Background: Diabetic gastroparesis is a disabling condition with no consistently effective treatment. In animals, ghrelin increases gastric emptying and reverses postoperative ileus. We present the results of a double blind, placebo controlled, crossover study of ghrelin in gastric emptying in patients with diabetic gastroparesis. Methods: Ten insulin requiring diabetic patients (fivemen, six type I) referred with symptoms indicative of gastroparesis received a two hour infusion of either ghrelin (5 pmol/kg/min) or saline on two occasions. Blood glucose was control led by euglycaemic clamp. Gastric emptying rate (GER) was calculated by real tim e ultrasound following a test meal. Blood was sampled for ghrelin, growth hormon e (GH), and pancreatic polypeptide (PP) levels. Cardiovagal neuropathy was asses sed using the Mayo Clinic composite autonomic severity score (range 0 (normal)- 3). Results: Baseline ghrelin levels were mean 445 (SEM 36) pmol/l. Ghrelin infu sion achieved a peak plasma level of 2786 (188) pmol/l at 90 minutes, correspond ing to a peak GH of 70.9 (19.8) pmol/l. Ghrelin increased gastric emptying in se ven of 10 patients (30 (6)%to 43 (5)%; p = 0.04). Impaired cardiovagal tone co rrelated inversely with peak postprandial PP values (p < 0.05) but did not corre late with GER. Conclusions: Ghrelin increases gastric emptying in patients with diabetic gastroparesis. This is independent of vagal tone. We propose that analo gues of ghrelin may represent a new class of prokinetic agents. Background: Diabetic gastroparesis is a disabling condition with no consistently effective treatment. In animals, ghrelin increases gastric emptying and reverses postoperative ileus. We present the results of a double blind, placebo controlled, crossover study of ghrelin in gastric emptying in patients with diabetic gastroparesis . Methods: Ten insulin requiring diabetic patients (fivemen, six type I) referred with symptoms indicative of gastroparesis received a two hour infusion of either ghrelin (5 pmol / kg / min) or saline on two occasions. Blood glucose was control led by euglycaemic blood was sampled for ghrelin, growth hormone (GH), and pancreatic polypeptide (PP) levels. Cardiovagal neuropathy was asses sed using the Mayo Clinic Composite autonomic severity score (range 0 (normal) -3). Results: Baseline ghrelin levels were mean 445 (SEM 36) pmol / l. Ghrelin infu sion achieved a peak pla sma level of 2786 (188) pmol / l at 90 minutes, correspond ing to a peak GH of 70.9 (19.8) pmol / l Ghrelin increased gastric emptying in se ven 10 patients (30 (6)% to 43 %; p = 0.04). Impaired cardiovagal tone co rrelated inversely with peak postprandial PP values ​​(p <0.05) but did not corre late with GER. Conclusions: Ghrelin increases gastric emptying in patients with diabetic gastroparesis. This is independent of vagal tone. We propose that analo gues of ghrelin may represent a new class of prokinetic agents.